Enhancing Loan Quality through Transparency: Evidence from ......Motivation Loan securitization is...
Transcript of Enhancing Loan Quality through Transparency: Evidence from ......Motivation Loan securitization is...
Enhancing Loan Quality through Transparency:
Evidence from the European Central Bank Loan
Level Reporting Initiative
Bank of France October 23, 2015 Aytekin Ertan, LBS
Maria Loumioti, USC
Regina Wittenberg-Moerman, USC
Research question
Can transparency alleviate agency costs in
loan securitization and improve loan quality?
Motivation
Loan securitization is an important credit market innovation that
allows banks to diversify credit risk and firms to gain easier
access to credit (e.g., Nadauld and Weisbach 2011, Benmelech,
Dlugosz and Ivashina 2012, Bord and Santos 2014, Loumioti and Vasvari
2015)
Securitization was blamed for playing a detrimental role in the
financial crisis by giving rise to severe agency problems in loan
underwriting, screening and monitoring (i.e., insufficient
screening and information collection [Keys et al. 2010, Garmaise
2015, Demyanyk and Hermert 2015], lower monitoring incentives [Kara
et al. 2011, Wang and Xia 2015]).
Motivation
In the aftermath of the crisis, market participants called for greater
transparency that would facilitate better assessment and pricing
of banks’ credit risk-taking.
However, whether transparency can effectively influence banks’
behavior and credit practices and alleviate risk-taking in
securitized transactions has yet to be empirically explored.
Setting
Starting from 2013, banks that pledge their ABS as collateral for
ECB repo borrowing are required to quarterly report to ECB the
structure and performance of the loans in these ABS deals.
Reporting is standardized and follows a predetermined format that
is set by ECB.
If a bank fails to report, it is barred from ECB repo borrowing.
Reporting and monitoring of the bank data is administered by an
independent entity, the European DataWarehouse (ED).
Hypothesis
Greater transparency will incentivize banks to issue and securitize
better quality loans.
Information collection
The comprehensive and recurring information collection required by
the new standards will result in a greater information set being
available to lenders when making credit decisions.
In turn, this will enhance banks’ screening efforts and underwriting
standards.
Market discipline
Detailed loan-level disclosure should assist investors to more
accurately assess the riskiness of securitized loan portfolios.
These disclosures are standardized and will therefore allow investors
to compare underwriting standards and securitized loan performance
across banks.
Confounding factors
Over the past few years the ECB – the primary investor in SME loan-
backed ABSs – has lowered credit standards in an effort to support
highly-leveraged banks (ECB Euro Money Survey [2012])
- Thus banks may not feel pressed to improve their securitized loan
quality.
The inherent complexity in securitized portfolio structures may deter
investors from effectively processing loan-level information.
- Investors expressed concerns with respect to integrating and
utilizing loan-level data to achieve efficient and consistent analysis
of ABS deals.
Greater information collection on loan and borrower performance might
not translate to better credit decisions by loan officers.
- To enhance credit decision quality, banks may also need to invest
in training personnel and improving their monitoring and control
systems.
Sample
974,717 loans to 606,396 SMEs issued by 37 banks from 2009 to
2014.
Loans are securitized in 73 ABS deals of banks in Portugal, Spain,
France, Belgium, Italy, Germany and The Netherlands.
We focus on small- and medium-sized enterprise (SME) loan
securitizations.
Sample selection
Observations
Securitized loans originated in 2009Q1-2014Q2 and reported in 2013Q1-
2014Q2 from 37 banks 3,845,770
Obscure/Codified data provider bank name 271,203
Missing interest rate, loan amount and/or maturity 448,527
Recording date is after loan maturity date or before loan start date 164,823
Total 2,961,217
Transparency and securitized loan quality
(I) (II) (III) (IV)
Default=1 Delinquent amount Number of days in
delinquency Loss given default
Transparency loan -0.020*** -0.195* -0.042** -0.024**
(-2.79) (-1.70) (-1.98) (-2.11)
Interest rate 0.011*** 0.305*** 0.041*** 0.004**
(9.41) (7.86) (7.34) (2.29)
Secured 0.017*** 0.193 0.051** -0.053*
(4.30) (1.05) (2.39) (-1.66)
Years to maturity -0.002 -1.405*** -0.007 -0.038***
(-0.68) (-5.41) (-0.56) (-4.83)
Securitized loan amount 0.001 0.816*** 0.063** 0.013
(1.14) (2.60) (2.10) (0.68)
Lending relationship -0.009*** -0.154*** -0.016*** -0.012*
(-3.20) (-2.61) (-2.58) (-1.79)
Loan purpose FE YES YES YES YES
Loan type FE YES YES YES YES
Borrower industry FE YES YES YES YES
Borrower type FE YES YES YES YES
Reporting quarter FE YES YES YES YES
ABS deal FE YES YES YES YES
N 2,729,323 2,961,217 2,961,217 2,961,217
Pseudo -R2 14.07%
Adj.-R2 4.43% 7.37% 44.21%
Transparency and securitized loan quality
Are the results driven by time –varying supply of
better loans?
Are the results driven by ECB monitoring of
banks?
(I) (II) (III) (IV)
Default=1 Delinquent amount Number of days
in delinquency Loss given default
Transparency loan -0.009*** -0.104* -0.029*** -0.003
(-3.28) (-1.92) (-3.14) (-0.70)
Interest rate 0.004*** 0.086*** 0.023*** 0.003
(6.04) (2.59) (3.26) (0.40)
Secured 0.003*** 0.029 0.003 -0.059
(2.49) (0.55) (0.46) (-1.16)
Years to maturity -0.005*** -0.313*** -0.020*** -0.046***
(-4.22) (-3.16) (-3.11) (-5.54)
Securitized loan amount 0.013 -0.835 0.048 0.090***
(0.43) (-1.24) (3.33) (3.59)
Lending relationship -0.004*** -0.066** -0.006 -0.019
(-2.61) (-2.04) (-0.74) (-1.18)
Loan purpose FE YES YES YES YES
Loan type FE YES YES YES YES
Borrower industry FE YES YES YES YES
Borrower type FE YES YES YES YES
Reporting quarter FE YES YES YES YES
ABS deal FE YES YES YES YES
N 161,216 167,985 167,985 167,985
Pseudo -R2 44.30%
Adj.-R2 7.55% 5.09% 24.67%
Transparency and securitized loan quality:
Loans originated in 2013Q1-Q2
(I) (II) (III) (IV)
Default=1 Delinquent amount Number of days in
delinquency
Loss given
default
Transparency loan -0.021*** -0.195** -0.043* -0.026**
(-2.68) (-1.96) (-1.87) (-2.00)
Interest rate 0.011*** 0.305*** 0.049*** 0.037**
(6.15) (7.86) (9.50) (2.17)
Secured 0.018*** 0.193 0.071** -0.041
(3.99) (1.05) (2.29) (-0.86)
Years to maturity -0.001 -1.405*** -0.016 -0.045***
(-0.43) (-5.41) (-1.09) (-4.09)
Securitized loan amount 0.012 0.816*** 0.093** 0.004
(1.00) (2.60) (2.45) (0.22)
Lending relationship -0.009*** -0.154*** -0.021*** -0.012
(-2.69) (-2.61) (-2.78) (-1.35)
Loan purpose FE YES YES YES YES
Loan type FE YES YES YES YES
Borrower industry FE YES YES YES YES
Borrower type FE YES YES YES YES
Reporting quarter FE YES YES YES YES
ABS deal FE YES YES YES YES
N 1,968,479 2,200,333 2,200,333 2,200,333
Pseudo -R2 15.21%
Adj.-R2
5.24% 7.27% 35.69%
Transparency and securitized loan quality:
Banks with ECB repo financing pre 2013
Transparency and securitized loan quality:
The role of information collection
(I) (II) (III) (IV)
Default Delinquent
amount
Number of days in
delinquency
Loss given
default
Transparency loan -0.027*** -0.188*** -0.054** -0.023**
(-2.75) (-2.86) (-2.45) (-2.31)
Information collection 0.414*** 1.503*** 3.428*** 0.901*
(3.37) (3.68) (2.81) (1.75)
Transparency
loan*Information
collection -0.265*** -1.605** -0.509*** -0.044
(-2.68) (-2.44) (-2.77) (-1.01)
Loan characteristics YES YES YES YES
Loan purpose FE YES YES YES YES
Loan type FE YES YES YES YES
Borrower industry FE YES YES YES YES
Borrower type FE YES YES YES YES
Reporting quarter FE YES YES YES YES
ABS deal FE YES YES YES YES
N 2,729,323 2,961,217 2,961,217 2,961,217
Pseudo-R2 15.43%
Adj.-R2
10.17% 8.23% 42.02%
(I) (II) (III) (IV)
Default=1 Delinquent amount Number of days in
delinquency
Loss given
default
Transparency loan -0.016*** -0.155** -0.020** -0.020**
(-3.28) (-2.27) (-1.99) (-2.18)
Peers’ transparency -0.078 -0.020 -0.027 0.007
(-1.45) (-0.43) (-0.57) (0.19)
Transparency loan*
Peers’ transparency -0.009** -0.021*** -0.010** -0.005
(-2.10) (-2.75) (-2.74) (-0.72)
Loan characteristics YES YES YES YES
Loan purpose FE YES YES YES YES
Loan type FE YES YES YES YES
Borrower industry FE YES YES YES YES
Borrower type FE YES YES YES YES
Reporting quarter FE YES YES YES YES
ABS deal FE YES YES YES YES
N 2,729,363 2,961,217 2,961,217 2,961,217
Pseudo-R2 13.83%
Adj.-R2
5.89% 7.47% 41.40%
Transparency and securitized loan quality:
Market discipline
Supplementary analyses
Do banks strategically select which ABSs to pledge to the
ECB as collateral for repo financing?
- A bank may opt to collateralize and report better performing
ABSs (i.e., ABSs secured by better quality SME loans),
while retaining low quality ABSs on its balance sheet.
Do banks sell worse quality ABSs to investors?
Can our findings be attributed to bank misreporting (banks
reporting stronger than actual loan performance)?
(I) (II)
Participating banks Matched sample
1-yr CDS
spread
5-yr CDS
spread
1-yr CDS
spread
5-yr CDS
spread
Post-transparency -0.005** -0.006** 0.003 0.004
(-2.32) (-1.78) (0.40) (0.76)
Reporting Bank*Post-transparency -0.005 -0.010**
(-1.32) (-2.33)
Controls YES YES YES YES
Quarter FE YES YES YES YES
Bank FE YES YES YES YES
N 152 152 311 311
Adj.-R2 81.23% 88.52% 81.70% 86.87%
Transparency and CDS spreads
(I) (II)
Participating banks Matched sample
Loan loss
provisions
Loan
impairments
Loan loss
provisions
Loan
impairments
Post-transparency -0.024** -0.017 0.001 0.010
(-0.60) (-1.10) (0.01) (0.40)
Reporting Bank*Post-transparency -0.060** -0.044**
(-2.01) (-1.72)
Controls YES YES YES YES
Quarter FE YES YES YES YES
Bank FE YES YES YES YES
N 280 280 560 560
Adj.-R2 53.60% 73.60% 66.40% 60.65%
Transparency and Bank’s financial performance
Transparency and prices of un-retained ABSs
(I) (II)
ABS credit rating
at origination ABS price
Post-transparency -1.009*** 0.017**
(-1.68) (2.34)
Controls YES YES
Bank FE YES NO
ABS quarter of origination FE YES YES
Trade quarter FE NO YES
ABS deal FE NO YES
N 155 735
Adj.-R2 72.44% 69.31%
Transparency and securitized loan quality
Are the results driven by EU-wide or national regulatory
changes in accounting for securitization risk and reporting
securitized transactions?
- CRD IV: more stringent requirements for banks to directly or indirectly
retain higher share of their securitized deals as well as for calculating
risk provisions in their securitized positions for capital adequacy
compliance purposes (effective date: Jan 1, 2014).
- Regulation 1060/2009/EC: credit rating agencies are required to
disclose more details of their standardized credit rating methodology
and processes (effective date: Sep 2009).
- Reg ECB/2008/30: banks engaging in securitization activities are
required to report to ECB various characteristics on their SPEs’ assets
and liabilities (effective date: 2010Q3) Spain adopted more
transparent reporting.
- Dutch securitization forum adopting standardized documentation for
securitized debt transactions (effective date: Jan 2013).
(I) (II) (III) (IV)
Default Delinquent
amount
Number of days in
delinquency Loss given default
Transparency loan -0.011*** -0.093 -0.029* -0.013***
(-3.48) (-1.40) (-1.82) (-3.01)
Loan
characteristics YES YES YES YES
Loan purpose FE YES YES YES YES
Loan type FE YES YES YES YES
Borrower industry
FE YES YES YES YES
Borrower type FE YES YES YES YES
Reporting quarter
FE YES YES YES YES
ABS deal FE YES YES YES YES
N 1,049,883 1,068,797 1,068,797 1,068,797
Pseudo-R2 23.71%
Adj.-R2 5.45% 7.20% 35.29%
Transparency, loan performance and the effect of regulatory
changes
Restrict sample to loans originated over the 2011Q1–2013Q3 period:
most major regulatory changes had taken place prior to the beginning
of this period and CRD IV went into effect after it.
Exclude loans by Spanish (most transparent national reporting
environment) and Dutch banks (concurrent regulatory change).
Transparency and securitized loan quality
Are the results driven by supervisory pressures and
international authorities’ scrutiny (e.g., EC, ECB, IMF)?
- Spain and Portugal have entered Economic and Financial
Assistance Programmes in 2011 and 2012.
- Italy also received rescue funds from the European Union (i.e., the
European Financial Stability Facility) in 2011.
- EC, IMF and the ECB have been launching regular review
missions to these countries to analyze economic, fiscal and
financial developments.
- Banks’ financial stability and long-term sustainability became
tightly linked to the rescue funds.
- The improvement in loan quality may be attributed to banks’
tightening their lending standards under greater uncertainty over
their countries’ long-term financial stability (e.g., Becker and
Ivashina [2014]).
(I) (II) (III) (IV)
Default=1 Delinquent amount Number of days
in delinquency Loss given default
Transparency loan -0.012*** -0.096** -0.035*** -0.019*
(-3.43) (-1.99) (-3.18) (-1.79)
Financial stability 0.003 0.042 0.011 0.002
(1.19) (1.38) (1.03) (0.48)
Transparency loan*
Financial stability -0.013** -0.011 -0.018** 0.005
(-2.24) (-1.52) (-1.98) (1.16)
Interest rate 0.009*** 0.162*** 0.037*** 0.006**
(7.26) (9.85) (6.11) (2.07)
Secured 0.011*** 0.087 0.031* -0.057
(2.64) (1.53) (1.72) (-1.36)
Years to maturity -0.001 -0.450*** -0.008 -0.042***
(-0.36) (-3.77) (-0.69) (-5.83)
Securitized loan amount 0.007 0.134 0.046** 0.049***
(1.42) (0.40) (2.02) (3.02)
Lending relationship -0.007*** -0.043 -0.010 -0.013
(-2.81) (-1.28) (-1.48) (-1.45)
Loan purpose FE YES YES YES YES
Loan type FE YES YES YES YES
Borrower industry FE YES YES YES YES
Borrower type FE YES YES YES YES
Reporting quarter FE YES YES YES YES
ABS deal FE YES YES YES YES
N 1,282,203 1,482,599 1,482,599 1,482,599
Pseudo -R2 18.13%
Adj.-R2 9.58% 7.23% 41.63%
Transparency, loan quality and the and the effect of
international monitoring pressure
Conclusions
Loans originated under the transparency regime are of better
quality in terms of their default probability, loss given borrower’s
default and late principal and interest payments.
Banks with better information collection as well as banks
operating in more transparent credit markets experienced greater
improvement in loan quality under the transparency regime.
The agency costs and risk taking inherent in
securitization can be alleviated by increasing
transparency.
Our results do not imply social welfare benefits, as the economic
benefits of transparency may not outweigh its costs (e.g., Dang et
al. 2014).
Contribution (1)
We add to the literature that explores the agency costs in loan
securitization (e.g., Keys et al. 2010, Keys et al. [2012], Benmelech
and Dlugosz [2009], Benmelech et al. [2012], Garmaise [2015]).
- By showing that higher transparency can alleviate inefficiencies
inherent in structured finance.
We contribute to the literature on the role of transparent reporting
in the banking industry (e.g., Beatty and Liao [2014], Granja
[2014], Bushman [2014], Acharya and Ryan [2015], Costello et al.
[2015]) and on the role of banks’ financial reporting quality in
alleviating risk-taking (e.g., Nier and Baumann [2006], Bushman
and Williams [2015])
- By documenting a direct link between reporting transparency via
loan-level disclosures and the quality of banks’ lending decisions.
Contribution (2)
We contribute to studies on the role of banks’ transparent
reporting in credit crises (e.g., Laux and Leuz [2010], Barth and
Landsman [2010], Laux [2012], Bischof and Daske [2013], Bischof
et al. [2015]).
- By stressing the role of transparency in alleviating risk-taking
behavior.
We extend we extend the emerging literature on the role of
transparency in improving operational decisions and
organizational performance (Christensen, Floyd and Maffett
[2014], Buell, Kim and Tsay [2015], Mohan, Buell and John [2015]).
- By delineating two important channels through which transparency
leads to better credit practices.
- We show that both the information collection and the market discipline
channels are instrumental in insuring higher loan quality.
THANK YOU